An important approach for extending depth-of-focus is to insert a phase mask directly into the imaging system to increase the focus invariance of the point spread functions (PSF). The process results in an intermediate image that is blurred in a spatially invariant way. In this paper we consider some linear and nonlinear iterative techniques for restoring imagery blurred by such wavefront coding. Tests are performed on data formed by artificially blurring images with an actual cubic phase mask PSF. From these tests we infer that all the methods produce adequate restoration, while a nonlinear, nonnegative constrained iterative algorithm due to Nagy and Strakos, called modified residual norm steepest descent (MRNSD), appears to produce especially accurate restoration for this problem, in an efficient and stable way.